def ret(input_img):
if not cutter.is_green(input_img):
return input_img
img=input_img.copy()
y1,y2,x1,x2=cutter.find_square_coords(input_img)
prediced=np.squeeze(self.model.predict(img),0)
img=np.squeeze(img,0)
img[y1:y2,x1:x2]=prediced[y1:y2,x1:x2]
return np.expand_dims(img,0)
def __init__(self,img_cols,img_rows):
"""
Initializes the autoencoder.
"""
self.set_training_info()
globals().update(self.info)
self.threshold=threshold
self.img_cols = img_cols#256 # Original is ~576
self.img_rows = img_rows#256 # Original is ~720
self.channels = 3 # RGB
self.img_shape=(self.img_cols,self.img_rows,self.channels)
dummy=plotload.load_one_img(self.img_shape, dest='med/green',extra_dim=True)
self.dims =cutter.find_square_coords(dummy)
self.model=None
self.pretrained=False
def mask_green_corner(imgs, val=0):
"""
Mask the bottom area where the green sqare is. takes a series of images (num images,image rows, image cols,channels) and masks an area with the chosen val
:param imgs: list of imgs
:param mask_width: width of the mask
:param mask_height: height of the mask
:param val: the value that the mask is filled with
:return: (The masked imgs,the part that is removed,pos of the removed part)
"""
img_num, img_cols, img_rows, img_channels = imgs.shape
y1, y2, x1, x2 = cutter.find_square_coords(imgs)
masked_imgs = np.empty_like(imgs)
missing_parts = np.empty((img_num, y2 - y1, x2 - x1, img_channels))
for i, img in enumerate(imgs):
masked_img = img.copy()
missing_parts[i] = masked_img[y1:y2, x1:x2, :].copy()
masked_img[y1:y2, x1:x2, :] = val
masked_imgs[i] = masked_img
return masked_imgs, missing_parts, (y1, y2, x1, x2)
def __init__(self, img_cols, img_rows):
"""
Initializes the autoencoder.
"""
self.set_training_info()
globals().update(self.info)
self.threshold = threshold
self.img_cols = img_cols # Original is ~576
self.img_rows = img_rows # Original is ~720
self.channels = 3 # RGB
self.img_shape = (self.img_cols, self.img_rows, self.channels)
if not mask:
dummy = plotload.load_polyp_batch(self.img_shape,
20,
data_type='med/stool-inclusions',
crop=False)
self.dims = cutter.find_square_coords(dummy)
self.combined = None
self.discriminator = None
self.generator = None
self.pretrained = False
def train_model(self):
"""
Trainer: Trains the loaded autoencoder model
:param epochs: number of epochs run
:param batch_size: how many imgs in each batch
:param save_interval: how many epochs between each save
"""
if self.info==None:
print("Warning no info found, prompting for info")
self.set_training_info()
globals().update(self.info)
if self.model==None:
print("Error: no model loaded")
return
if self.pretrained==True:
print("Warning: model has pretrained weights")
from tqdm import tqdm
y1,y2,x1,x2=cutter.find_square_coords(plotload.load_polyp_batch(self.img_shape, batch_size))
for epoch in tqdm(range(epochs)):
X_train=plotload.load_polyp_batch(self.img_shape, batch_size,data_type='med/none')
if mask==0:
Y_train,X_train=cutter.add_green_suare(X_train)
else:
print("Not yet implimented")
cur_loss=self.model.train_on_batch(X_train, Y_train)
if epoch%10==0:
self.save_img(epoch)
if cur_loss<self.threshold:
print(cur_loss)
self.threshold=cur_loss
self.model.save(f"models/AE-{self.img_shape[0]}-{self.img_shape[1]}-{'c' if mask==0 else 'n'}.h5")
self.model.save_weights(f"models/AE-{self.img_shape[0]}-{self.img_shape[1]}-{'c' if mask==0 else 'n'}-w.h5")
self.model.save(f"models/AE-{self.img_shape[0]}-{self.img_shape[1]}-{'c' if mask==0 else 'n'}-fin.h5")